Why Your Brand Research Is Lying to You

Imagine your startup has interviewed twenty target buyers about a new premium tier. Every participant claims they would pay a forty percent markup for the added security features you described. You are now deciding whether to commit your remaining quarterly budget to a high-end brand pivot based on these encouraging signals. Is this evidence a green light or a trap?

The Gap Between Sentiment and Action

Interviews are effective for exploring general attitudes and how people conceptualise a problem. They allow a founder to hear first-hand stories about frustrations and needs that might not appear in raw analytics. A limitation of the format is its reliance on participants remembering past use or imagining future behaviour. Neither response, on its own, establishes what a buyer will purchase. A retrospective account can make past actions appear more consistent than the evidence warrants. Treat it as an account to investigate, rather than a complete record. [2] [5]

This tendency to interpret information as part of a larger, coherent story is known as narrative bias. When a founder asks why a buyer chose a competitor, the answer might emphasise price or features. Treat that explanation as a hypothesis. Check it against the options, circumstances and constraints present when the decision happened. In my interpretation of brand strategy, relying on these rationalisations leads to positioning built on what people think they should want. You risk solving a perceived problem that does not exist in the wild. [4] [5]

The Query Effect and Stated Preferences

The very act of asking a question can distort the data through the query effect. People can form an opinion about almost anything if prompted. This happens even if they had never considered the topic before. If you ask about your visual identity, you may collect feedback on colours and fonts. This does not mean those elements are important to their buying decision. You have evidence of an opinion expressed in that conversation. You still need to establish whether it affects choice. Keep feedback separate from evidence of a commitment to act. [5]

Describing an unfamiliar feature asks participants to imagine a future situation. Their answers cannot confirm what they will do once the feature exists. This makes 'would you use this' a dangerous question for a founder. In a strategic context, these stated preferences should be treated as evidence of current mental models. They reveal how a person talks about a category. This is useful for messaging but insufficient for validating a pivot. [5] [6]

Observing the Trail of Real Decisions

To examine a positioning hypothesis, combine retrospective accounts with observations from situations you can investigate. Field studies and direct observation allow you to see what users actually do in their natural environment. For example, a customer-service representative might report following a specific protocol during an interview. Observation might reveal they actually use a series of convoluted workarounds. These workarounds can suggest a problem worth investigating. They do not establish a market or a price. Ask how they affect effort, time or the choice of a supplier. [2]

Diary studies offer another way to reduce the errors of retrospective recall. By asking participants to log actions and feelings in real-time over days or weeks, you capture data closer to the moment of experience. Diary studies reduce reliance on retrospective recall. They do not eliminate memory errors, self-report bias or participant bias. From a brand strategy perspective, this longitudinal data reveals the actual frequency of a problem. This helps a founder distinguish between a one-off annoyance and a persistent pain point worth a premium solution. [2]

Triangulating Conflicting Evidence

When interviews suggest one direction and observation suggests another, the conflict is a reason to investigate. Ask whether both methods examine the same audience and situation. Differences in timing, participants or measurement may explain the conflict. If people request a feature but no related action is observed, you have two different signals. That does not prove either is false. Explore constraints the interview may have missed. Agreement across methods can strengthen an explanation. It cannot rescue a poorly defined question. Keep the decision you are investigating explicit when combining interviews, analytics and field research. [3] [5]

Before committing your budget, verify the base rate of the behaviour you are targeting. One vivid interview should not supply your estimate of market size. A story about a buyer and an estimate of how many buyers exist answer different questions. If your positioning relies on a highly specific set of user traits, check if that segment is large enough to sustain the business. My strategic recommendation is to use interviews to generate hypotheses. Use observational data or small-scale live tests to validate them. Positioning is too expensive to build on polite conversation. [3] [4]